Normal CSPMEI Circuit Reference Atlas — Enhancers, Regulome, Promoters & Coding Gene regions with TF binding sites (hg38)
The Single-Cell Genomic Regulatory Atlas and Map (scGRAM) is an interactive genome browser for exploring transcription factor binding landscapes across single cells. Built on the hg38 human reference genome, it enables researchers to visualize and compare regulatory elements at single-cell resolution.
Visualize regulatory regions and TF binding sites as interactive tracks on the hg38 genome.
Click any region to see a detailed SVG visualization of TFs bound to sense and antisense strands.
Overlay up to 3 cells side-by-side to compare regulatory profiles.
View the same cell's data across all regulatory element types simultaneously.
Circos plot, TF-Gene network, and Cis-Trans map showing the complete regulatory architecture of a cell.
Aggregate visualization across all cells in a subtype to reveal population-level regulatory patterns.
Project cells onto a 2D map based on TF footprint presence in a chosen region. Reveals subpopulations separated by regulatory activity at a specific locus.
Search by gene name, TF motif, or genomic coordinates (chr1:1000000-2000000).
Download regions and TF binding sites as BED files for downstream analysis.
scGRAM (Single-Cell Genomic Regulatory Atlas and Map) provides a comprehensive view of transcription factor binding across single cells from the Normal CSPMEI Circuit Reference Atlas.
The data comprises single-cell level regulatory element annotations including enhancers, regulome elements (silencers, CAGE clusters, DNase I hypersensitive sites), promoters, coding gene regions, and open chromatin peaks derived from Genrich peak calling. TF binding site predictions are mapped to each regulatory element.
TF binding sites were identified by overlapping regulatory regions with motif databases. Enhancer annotations include GeneHancer-derived regions. Regulome elements encompass multiple regulatory categories. Promoter regions are defined by upstream regulatory windows with associated TF footprints. Open chromatin peaks were called using Genrich on single-cell BAM files.
Built with Flask (Python), IGV.js, D3.js, and Chart.js. The genome browser uses the hg38 human reference assembly. All static assets are served locally for reliability.
If you use scGRAM in your research, please cite: [Citation information to be added]
For questions or feedback, please contact the Guda Lab.
1. Go to the Browser tab. 2. Select a Data Type (Enhancers, Regulome, Promoters, Coding Gene, or Open Chromatin Peaks). 3. Choose a Dataset, Subtype, and Cell. 4. Click Load Tracks to visualize in the genome browser.
After loading tracks, click any row in the Regions table to open the TF Binding Landscape. This shows an unwound DNA view with TFs positioned on the sense (+) and antisense (-) strands.
Use the search box to find regions by: Gene name (e.g., BRCA1), TF motif (e.g., CTCF), Genomic coordinates (e.g., chr1:1000000-2000000), or Regulome entity type (e.g., enhancer, silencer, CAGE_cluster, promoter, protein_bind, locus_control_region, DNase_I_hypersensitive_site, rep_origin, enhancer_blocking_element, and more). The search uses substring matching, so partial terms work (e.g., "protein_bind" matches all protein binding entries).
Click + Compare to add the current cell to the comparison list (up to 3). IGV will overlay tracks from all selected cells.
Click Cross-Type View to see the same cell's barcode matched across all available data types (enhancer, regulome, promoter, gene).
Click Cell Regulatory Map for a full-page visualization including a Circos plot, TF-Gene network, and the Cis-Trans Regulatory Map showing DNA elements connected by shared gene targets.
Click Subtype Landscape (available when a subtype is selected) for a population-level view showing which regulatory regions are shared across cells, the most frequent TFs, and gene hub connectivity.
Use Cluster by Region in the Browser action bar to project all cells in the current subtype onto a 2D map, or open the standalone Clustering tab from the top navigation. For each cell, a binary feature vector indicating the presence of each TF within the chosen region is built; UMAP (default) or t-SNE is then applied. Points are colored by Has Footprint (binary), Density (tiered by unique TF count), or Dominant TF. A region must have TF footprints in ≥2 cells to be eligible. Hover points for per-cell details; scroll to zoom, drag to pan. Export as PNG or CSV. Note: Open Chromatin Peaks are not supported as they do not contain TFBS.
Use Export Regions or Export TFBS to download data as BED files compatible with UCSC Genome Browser and other tools.
Project all cells of a subtype onto a 2D map using tSNE or UMAP based on the transcription factor footprints present within a chosen genomic region. Cells are colored by whether they have any footprint in the region ("has footprint" vs "no footprint"), with optional sub-colorings by footprint density or dominant TF. This reveals how cells split based on regulatory activity at a specific locus.
Tip: Only regions with TF footprints in ≥2 cells of the subtype will produce a meaningful embedding. Peaks data type is not supported (no TFBS).
Cluster cells using TF footprints across multiple regions from any combination of data types. Cells are matched by barcode across types. Each region contributes its own TF feature space — the embedding reflects the combined regulatory signature across all selected loci. Up to 50 regions.
| # | Chromosome | Start | End | Size (bp) | Gene | #TFs |
|---|
Cluster all cells in a subtype using TF footprints across multiple regions from any combination of data types. Up to 50 regions.